Slashdot Mirror
2nd Multi-Format 128kbps Public Listening Test
technology is sexy writes "Roberto Amorim has launched his latest public listening test evaluating the performance of different audio codecs at 128kbps, among them Apple's AAC implementation (used in iTunes), LAME, Ogg Vorbis fork auTuV, WMA, Musepack and even Sony's Atrac3 format, which is soon to be used in their own music store. Read more on Hydrogenaudio and check out the results of prior tests. As opposed to most evaluations of audio codecs, this is a scientific test adhering to ITU-R BS.1116-1 as much as possible while still allowing everybody to participate."
Here's insightful. Ogg is a wrapper. It has nothing to do with the quality of the sound. You should be chanting Vorbis.
Etiquette is etiquette. He kills his mother but he can't wear grey trousers.
I know you can do frequency analysis on the output of these various codecs. Just compare that to the average human auditory capacity and you can get an objective measurement of the merits of these various compression methods.
So uh, why is this necessary, exactly?
Of course, if that turns out to be inferior to any of the other formats, it would prove that something's wrong with the tests.
umm, no. "128kbps" = 128 kilo bits per second. .wav's sound the best, bitrate not restricted, but if you could make a wav 128kbps it would sound like utter .. crap
Different codecs and implementations of those codecs may be optimized for different bitrates, so its important to test codecs at various target bitrates.
DON'T CLICK THE LINK!
The sad thing is that somebody went to the trouble of putting together a perfectly reasonable, logical post just to throw in a porn link. *sigh*
Karma: Segmentation fault (tried to dereference a null post)
Just because you don't have a use for 64k audio, doesn't mean the results are meaningless. Lots of people have small-capacity players, and some codecs can tolerate that bitrate for very casual listening (such as in the car). Lots of streaming audio sources are at this bitrate or lower. Satellite radio is at 64k or lower. Also, it's not a good idea to try to extend these results to other bitrates. MPC for example, isn't even worth considering at 64kbps, but at bitrates over about 140kbps, it will beat the pants off of anything else.
A .wav file at 128kbps is going to sound absolutely awful. At 8 bits per sample (which sounds pretty bad no matter what), 128kbps gives you a sample rate of only 16khz, so any frequencies above 8khz will be lost. If you up the sample quality to 16 bit (CD quality), the sample rate goes down to 8khz (4khz frequencies).
And this is for monaural sound. If you want stereo, cut the sampling rate in half -- this might cut it for voice, but it won't work for anything else.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
There are music stores that sell music in 192kbps..For ex Real's music store..FOr anyone jumping out with a gun to announce the dismal performence of real audio, hold on...it uses AAC..Here is the info from Real's music store:
RealPlayer Music Store tracks are encoded in secure RealAudio 10 (.rax) format with Advanced Audio Coding (AAC) at 192Kbps. With AAC, RealPlayer Music Store offers the best-quality music downloads on the Web.
AAC is widely accepted as one of the highest quality audio formats for distributing digital music on the Internet. When compared to the popular MP3 format, AAC offers higher quality audio reproduction at lower bit rates.
-----------
And for all the posts that are going to trash Realplayer for adware, quit whining and start downloading RealPlayer 10 with all stupid features removed..
I must be deaf, I just did the test on a the kraftwerk sample file, and it took me a lot of relistening to finally pick out 3 out of 6 encoded files (although the first one - whatever it was - was fairly easy). The other 3 sounded exactly like the reference sample to me. This is using Sennheiser HD500 headphones and an Audigy ZX2 sound card.
Try doing the test, you might be surprised, or conversely if you're not surprised, you might contribute valuable information to the project.
Switch back to Slashdot's D1 system.
You are 100% clueless, pardon my french.
.wav file is about 1.5Mbps.
The bit rate of
Vorbis does variable bit rate and you set the quality you want. That way you don't waste lots of bits where they are not needed. My 4MB ogg file sounds as good or better than my little brother's 6MB mp3. The difference is more songs on my 256MB compact flash card. Yes, it's easy to play that music on my Zaurus, which cost about as much or less than DRM gimped portable music players.
I hate it when I download music (legal ;) and the only bitrate available for the song i want is 128.
Cry me a river.
Friends don't help friends install M$ junk.
The r3mix tuning (--r3mix), while a small step forward, was inherently flawed because of his insistance on tuning based on pictures instead of acual listening tests. As a result, the --dm-presets were invented and improved by Dibrom (the HydrogenAudio founder) along with a multitude of testers. eventually those were included in LAME as the --alt-presets (and in the latest version they just replace the normal --presets). In short, Hydrogen Audio is THE place to go for this stuff now.
Jeremy
That was my first reaction, who uses 128. What I want is a blind test with experts and thousand dollar audio systems to find at what point the experts are no longer able to tell the difference between the compressed and uncompressed audio.
I use `lame --preset standard`, which ends up being VBR in with a max of 110-290, hovering mostly around 190-210 range. It's one of the reasons I don't use OGG, it doesn't have any preset's so I'm supposed to just decide on a good level myself. I'd rather use something that it appears someone has put some thought into.
While the sine wave's frequency is known exactly (within the resolution of your sampling frequency) the amplitude is not- you always have loss due to quantization noise. You may be thinking of the fact that the fourier transform will have only one harmonic and thus the quantization noise doesn't come into play.
Consider the signal to quantization noise rate (SQNR):
SQNR (dB) = 20log(Vsignal/Vquantization_noise)
With linear quantization, your quantization is evenly spaced and the noise is 0.5, with a range of -2^(n-1) to +2^(n-1) and a single bit gives roughly 6dB of resolution.
Read these threads. Preparing Vorbis for the next multiformat test Vorbis Listening Test
People are constantly comparing audio coding standards, but realize that most of the stuff you hear is marketing speak. Many companies have lots of IP in this area and they obviously want to make their solution the standard.
:) Those books can provide a good basis of how all the coders tested works.
What makes one codec sound different than others is the psychoacoustic measures implemented, quantization method, and the windowing scheme implemented before MDCT is performed. Note that all of the coders tested there do not use the same windowing method, but all of them use MDCT in a way.
MP3 is a subband coding, it slices the audio into sub bands before transforming them. AAC, OTOH, is not. AAC uses straight MDCT and does the filtering there. The criteria for filtering is still the same old, tho. That is part of the reason why AAC at 128 kbps way outperform MP3.
Psychoacoustic is not new, it's been described extensively in a book "Psychoacoustic" by Fastl. The catch is, audio coders have to take into account the complexity of performing the full model. MP3 uses a very simplified version of it, and it taxes the highest spec of its day. That is also the reason why AAC-LC (low complexity) is more popular than AAC-Main profile nowadays.
Vorbis can sound better because with new hardware, a more mathematical heavy version of psychoacoustic can be implemented today. Plus, they discard the notion of constant bitrate and use quantization quality instead. This is also evident in FAAC.
128 kbps stereo is practically the limit of almost-transparent quality audio now. 64 kbps mp3pro is just bull, it doesn't perform anywhere close to modern mp3 at 128 kbps. There is a limit on compression, and that is governed by the entropy (information content) of a signal. You go lower than entropy, you lose information, simple as that. Having said that, the only way to reduce entropy is using psychoacoustic models, and that also have a limit.
Note also that Dolby-AC3 that is used in DVD and movie theatre compresses 5.1 channels into 384 kbps, or roughly 150-ish kbps stereo. Again, the same lower limit is evident. They do compression by combining the high frequencies > 15 khz and ignore the phase information in that high frequencies. As you can probably tell, AC-3 sounds pretty good.
If you're interested in this area, I suggest the MPEG-4 book by Ebrahimi, Psychoacoustic by Fastl, Multimedia Compression by Gibson and DSP First by I forgot who